After last’s week post calculating the success odds for the Lucky feat and Halfling racial ability, some folks who run the Lucky feat a little differently asked me to run the odds on that iteration. Since this homerule is popularly employed, including by D&D Lead Rules Designer Jeremy Crawford, it warrants taking another look.

RAW vs. RAI

D&D Lead Rules designer Jeremy Crawford confirmed that the Lucky feat can be more powerful with disadvantage in the 2016 Sage Advice Compendium Update (v 1.09). In a later appearance on the Dragon+ podcast, he explains further:

What happens if there’s advantage or disadvantage on the roll and you decide to use a luck point?

I made a ruling in the Sage Advice column which would be considered a RAW ruling. Based on the text of the feat, it says you spend the luck point, then you roll a die, and you get to choose what die is used in the end. My ruling was, based on the text we published, if you have advantage or disadvantage, it still works this way: you pick the one you use…

The Lucky feat is agnostic when it comes to advantage and disadvantage… What that means is in the context where you have advantage and disadvantage, you roll an additional dice. You have 3 dice you now get to choose from.

– Lead Rules Designer Jeremy Crawford, December 2016 Sage Advice podcast (@ 4:00)

The alternative interpretation plays on the rules-as-intended (RAI), instead of the strict rules-as-written (RAW) we examined in last’s week’s post.

Jeremy Crawford explains how he runs the Lucky feat at his own table, which is closer to the rule the design team intended:

This is where we get into the difference between rules as written and rules as intended. Really, the way I run this feat at my table is… I have advantage or disadvantage kick in on the original roll, then have the Lucky feat applied.

So, the way I would typically run it is: The person rolls the d20, and let’s say they have disadvantage. So they roll the second d20. They take the lower of those two dice. Then they spend their luck point, they roll a d20, and they choose which of those two to use.

So they’re still getting to do exactly what the feat said, which is pick which one to use. But they’re picking after disadvantage has been applied to the original roll.

– Lead Rules Designer Jeremy Crawford, December 2016 Sage Advice podcast (@ 10:15)

Although he explains that the Lucky feat is not demonstrably overpowered in play, contrary popular sentiment has led a lot of players to run RAI at their tables. In fairness to those players, we’ll run the numbers on RAI to see what they’re getting out of it. We can also compare those odds against RAW to see how much of a change the homerule makes.

RAI Odds

You can see that you’re still getting a pretty hefty bonus from the RAI Lucky feat, even when rolling with disadvantage.

Of course, RAI is less impactful than running it RAW, as you can see from this chart which shows the difference:

The average difference across all rolls is about 16.7% (41.6% RAW vs. 24.9% RAI), with the difference generally becoming more pronounced the higher the roll needed.